會計師核閱與揭露過去財務預測準確性對投資人是否信賴公司財務預測的影響

Abstract

對於公司之前景，公司與投資者間有資訊不對稱的問題，透過公司發佈財務預測乃是減緩此種資訊不對稱的方式之一。我國財務預測之相關規定，要求公司發布財務預測前，需經會計師核閱。研究乃是探討經會計師核閱之公司財務預測，是否會提升投資人對公司財務預測之信賴程度。 另外，本研究也探討公司過去財測準確性（含高估、小幅誤差與低估）對投資人給予公司財測信賴程度之影響。 本研究使用實驗法探討上列問題，本研究以某大學管理碩士學分班93位之學員作為受試者，以代表一般投資人。結果發現投資人在無法得知公司過去財務預測達成率的情形下，當期公司的財務預測經會計師核月後，會增加投資者對財務預測的信賴，減少調整的幅度。當投資者獲得過去財測達成率的資訊時，對公司未來獲利的預估會依據財務預測以及其過去的達成率向下或向上做調整，而財測是否經會計師核閱則無顯著影響。

This thesis examines whether (1) auditor’s review of management forecast and (2) historical record of management forecast accuracy will affect investors’ reliance on management forecast. Adopting a laboratory experimental approach, this study involves presence of auditor’s review and accuracy of past management forecast (both are between-subjects factors) and presence of historical record of management forecast accuracy (within-subject factor) as the independent variables, and investor’s reliance on management forecast as the dependent variable. Accuracy of past management forecast is manipulated at three levels, including accurate, overestimated, and underestimated, and investor’s reliance on management forecast is measured by the ratio of investor’s belief in earnings (and gross margin) after reading management forecast to management forecast of earnings (and gross margin).

The subjects in this experiment are 93 practitioners who attend an executive development program at a university. They are randomly assigned to the experimental conditions as mentioned above and serve as the role of investors. The results indicate that when investors are not provided with past record of management forecast accuracy, presence of auditor’s review of management forecast will increase investor’s reliance on management forecast. When investors are provided with past record of management forecast, the review function becomes insignificant in affecting investor’s reliance on management forecast. Accuracy of past management forecast significantly affects investor’s reliance on management forecast. Investors adjust management forecast downward when past forecast is overestimated, and adjust upward when past management forecast is underestimated with the magnitude of the former greater than the latter. The findings may have implications for the government in setting policies regulating management forecast. Limitations of this study are also discussed.